Potential Effects of Congressional Welfare Reform Legislation on Family Incomes Sheila R

Potential Effects of Congressional Welfare Reform Legislation on Family Incomes Sheila R. Zedlewski, Sandra J. Clark, Eric Meier, Keith Watson

The opinions expressed herein are those of the author and are not necessarily those of the Urban Institute, its trustees, or its sponsors. Document date: July 26, 1996 Released online: July 26, 1996 Table of Contents

Introduction and Summary Effects of Current Social Welfare Spending Proposed Changes to Social Welfare Spending Estimating Effects on Families Effects on Program Spending Effects on Poverty Effects on Total Family Incomes Sensitivity of Results to Assumptions Some Possible Modifications to the Legislation Conclusions Methods and Data (Appendix) LIST OF TABLES Table 1: Summary of Major Changes in Income Security Programs Proposed by H.R. 3734 Table 2: Budget and Caseload Effects of Major Provisions of the Welfare Reform Act (H.R. 3734) Table 3: Effects of H.R. 3734 on Total Family Income and Poverty Table 4: Characteristics of Families Losing Income Under H.R. 3734 Table 5: Effects of H.R. 3734 on Poverty Using Alternative Labor Supply Assumptions Table 6: Effects of Modifications to H.R. 3734 on Poverty and the Poverty Gap

The authors would like to thank Paul Johnson for outstanding programming assistance and Isabel Sawhill for helpful comments on an earlier draft. The opinions expressed herein are those of the authors and do not represent the views of the Urban Institute or its sponsors.

Introduction and Summary The House and Senate both recently passed sweeping welfare reform bills that would substantially change the current social safety net. These bills would replace Aid to Families with Dependent Children (AFDC) with block grants to the states with a fixed federal payment. In addition, states would be prohibited from using federal money to provide assistance to families for more than five years. The bills also would eliminate federal welfare payments to legal immigrants, reduce SSI benefits for children, and enact a variety of changes in the Food Stamp Program. The welfare reform legislation is designed to decrease dependency on government assistance and to shift more responsibility for social programs to states. Proponents argue that the changes will improve administrative efficiency and increase work and earnings by welfare recipients, thereby saving public funds with a minimal change in material hardship and poverty. Critics argue that the magnitude of the changes in welfare spending is so large that low-income families would be hurt and poverty increased, even if the changes motivate more welfare recipients to exchange dependency for employment. This paper presents estimates of how the major income security changes proposed by H.R. 3734 would affect family incomes.1 We do not include a number of other changes which would have smaller effects on family incomes nor do we consider provisions affecting Medicaid eligibility. Income is broadly defined in this analysis to include cash income; "near cash" income from food stamps, energy and housing assistance, and the earned income tax credit; less taxes. Our methods and analysis follow those used in an earlier study that examined the effects of the 1995 Budget Reconciliation Act on family incomes (Zedlewski, Clark, Meier, and Watson, 1995).2 We estimate the effects of the income security and tax changes in the welfare reform bill by comparing how families will fare under new rules compared to current program rules, making reasonable assumptions about how program changes will affect behavior. Since no one can predict these responses with certainty, we also show the sensitivity of the results to alternative behavioral assumptions. Our findings show that the proposed welfare reform changes would increase poverty and reduce incomes of families in the lowest income group. With the legislation fully phased in, spending on the current social safety net would be reduced by about $16 billion per year compared to current law. We estimate that 2.6 million more persons would fall below the poverty line as a result, including 1.1 million children. More than one-fifth of all families with children would see their incomes fall by about $1,300 per year, on average. Almost half of the families adversely affected by the legislation currently work, and 4 out of 5 families who would be affected currently have incomes below 150 percent of poverty. Alternative assumptions about how many welfare recipients move into jobs when their benefits are eliminated do not produce dramatically different results. Most families on welfare are already poor, and most long-term recipients who leave welfare for jobs will not earn enough to move above the poverty line. The proposed changes in the AFDC, SSI, and Food Stamp Programs would reduce the incomes of a large group of low-income families. We show that the effects of the legislation could be mitigated by allowing states to continue assistance for children whose parents do not successfully find work after five years; by increasing the percent of the caseload states can exempt from the time limit; by restoring eligibility for legal immigrants, and by restoring current food stamp benefit provisions for citizens. Effects of Current Social Welfare Spending The federal government currently spends about 7 percent of its budget for income security programs that are designed to help low-income families. About $107 billion of the federal budget goes directly to families to fund cash assistance (Aid to Families with Dependent Children, AFDC, and Supplemental Security Income, SSI), food stamps, housing, and energy assistance, child nutrition programs, and the refundable Earned Income Tax Credit (EITC). State governments spend an additional $17.5 billion to help finance income security programs.3 Three-quarters of the assistance under these programs is provided to families with children. A large group of American families depends to some extent on the nation's social welfare system.4 Approximately one in four families receives assistance from at least one of the programs listed above. Among low-income families with incomes below 130 percent of poverty, government programs assist 6 out of 10 families. These benefits account for nearly one-third of the total income of this group. Most of the rest of their income comes from earnings. Consequently, these income security programs substantially reduce poverty, especially for families with children. When poverty is measured so that income includes not only all sources of cash but also noncash government support, the number of persons who are poor drops from 43 million to 28 million. The poverty rate for children also drops from 24 percent to 14 percent. Proposed Changes to Social Welfare Spending The welfare reform bill would make major changes to the three largest social welfare programs — Aid to Families with Dependent Children (AFDC), Supplemental Security Income (SSI), and the Food Stamp Program. The major changes proposed in the bill are summarized in Table 1.5 The most fundamental change proposed by the House welfare reform bill would transform the AFDC program from an entitlement to a lump-sum payment to states for Temporary Assistance for Needy Families. The bill also would prohibit states from using the block grant funds to provide assistance to families who have received benefits for five years. This marks a major shift in the AFDC program from an indefinite source of support to a temporary and transitional one. States could exempt up to 20 percent of their caseload from the five-year time limit based on state-defined criteria. Two other changes would eliminate the $50 child support pass-through for AFDC families and require the state to implement a "family cap" unless the state legislature enacts legislation to opt out of this provision. Most noncitizens would lose eligibility for federal assistance under the proposal. AFDC benefits would be denied to noncitizens for their first five years in this country and states would have the option to categorically deny them benefits at any time. Noncitizens would also lose eligibility for SSI and food stamps. In addition to the noncitizen provisions, the major change proposed to the SSI program would eliminate benefits for disabled children who meet the current law eligibility criteria based on Individual Functional Assessments (IFAs) or maladaptive behavior. Food Stamp program rules would be modified to increase countable income, reduce benefits, require older children to file with their parents, and limit nonworking childless adults to a one-time, three-month period of eligibility.6 The legislation would also enact a means-test for meals served to children in Family Day Care Homes under the Child and Adult Care Food Program. We also assume that child support collections for parents receiving IV-D services would increase due to the bill's provisions to improve child support enforcement.7 Estimating Effects on Families In addition to the direct effects of the program rule changes outlined above, the proposed policies would lead to changes in behavior for states and families. The magnitude and complexity of the full package of proposals make it hazardous to predict how states and families will respond. But such behavioral responses may be critical in determining the overall impact of the policy changes on family incomes and government spending; it is essential to make any assumptions about such responses explicit. We use a fairly optimistic scenario for our basic estimates. First, we assume that all states will maintain their current benefit payments, reducing benefit levels only to compensate for the decreased federal contribution. Second, we assume that all states exempt the full 20 percent of their caseload allowable under Federal law. Based on research on how individuals respond to changes in financial incentives, we predict that two-thirds of mothers who lose welfare eligibility as a result of time limits will find jobs, most at part-time hours.8 We assume that they find year-round jobs with an hourly wage of nearly $6.00.9 Most mothers would try to work part time to maximize their net incomes, after work-related expenses and eligibility for other social assistance are taken into account. Whether the labor market could provide jobs for all these mothers, especially considering their low skill and education levels is uncertain.10 11 To allow for this uncertainty we show two alternative scenarios. One considers the possibility that fewer people find permanent jobs, using an estimate that 40 percent of long-term welfare recipients will find work from an analysis of earlier reform proposals completed by the U. S. Department of Health and Human Services and the Office of Management and Budget.12 This scenario assumes that 40 percent of long-term welfare recipients not exempted from the time limit would find work. The other, more optimistic scenario, assumes that two-thirds of long-term recipients will find jobs, half at full-time hours. Effects on Program Spending When fully implemented, the welfare reform bill's provisions would reduce government spending on the major welfare programs by nearly $16 billion annually compared to current law. Full implementation of the bill is scheduled to occur in 2002 when the five-year AFDC time limit would take place. Table 2 summarizes the effects of the proposed policies at full implementation. This is equivalent to simulating the new policies as if they were implemented in 1996. AFDC program spending is reduced by $5.3 billion, with nearly one million families losing all AFDC assistance. Spending on the SSI program falls by nearly $5 billion (20 percent), and nearly 900,000 fewer persons receive benefits. The bill reduces food stamp spending by $5.5 billion (23 percent) overall and eliminates 850,000 families from the program, but these effects reflect several sources of change, in addition to the program rule changes.13 Effects on Poverty We estimate that the welfare reform provisions would increase the number of persons in poverty by 2.6 million (Table 3). This estimate uses a measure of poverty that includes families' cash income and noncash income from food stamps, housing assistance, and the EITC, less taxes.14 Since the bill makes a number of changes in social welfare programs that affect both cash and noncash benefits, this broader measure of poverty more accurately captures the effects of the proposed changes. cash and noncash benefits, this broader measure of poverty more accurately captures the effects of the proposed changes. Most of the increase in poverty occurs for families with children. About 1.1 million more children would be moved into poverty, an increase of about 12 percent. In addition, the poverty gap — the difference between poor families' incomes and the poverty threshold — would increase by around $6 billion for all families, a 12 percent increase relative to current policies. Effects on Total Family Incomes About 10 percent of all families would lose income as a result of the proposed legislation (Table 3). Three-quarters of the families losing income have children. This is to be expected since many of the assistance programs affected by the bill target families with children. Eight million families with children would sustain average income losses of about $1,300 per family. The characteristics of families who would lose income under legislation vary by family structure, employment status, and income. Table 4 shows that three-quarters of these families have children; 45 percent are in single-parent families and nearly 30 percent in two-parent families. More than half are working families. Our results also show that more than three-quarters of these families currently have incomes below 150 percent of poverty. Thus, the legislation would reduce the incomes of many different types of families, including many who do not rely on cash welfare payments. Income losses are smallest on average for families who work and do not receive cash assistance, but even these families lose almost $1,000 on average (see Table 4). In contrast, two-parent families with children and families combining work and cash assistance would lose over $1,500 on average. The largest income losses, however, would be incurred by families above 150 percent of poverty, primarily because many of these families receive high SSI benefits or assistance from several programs. Sensitivity of Results to Assumptions The cumulative effects of the bill's provisions would result in major shifts in government assistance to families. Because we cannot know for certain how families and states would respond to the new social legislation, we provide additional alternative estimates of the effects of H.R. 3734 under two alternative labor market outcomes. The first more pessimistic alternative assumes that only 40 percent of the nonexempt caseload facing time limits will find jobs. The second more optimistic alternative assumes that two-thirds would find jobs, and half of those mothers would work full time. We do not make alternative assumptions about state behavior. States would have more flexibility but far fewer federal dollars. They also would have a new and potentially costly mandate to move their welfare population quickly into jobs. Either they would be forced to reduce eligibility and benefits for AFDC families (for example, by setting shorter time limits), or they would need to replace lost federal funds with their own resources. We have no way of predicting which states would add their own resources to the federal-state pot, or which would spend less of their own money on assistance to the poor. Therefore, we maintain the assumption that states would not reduce welfare spending for all of our estimates. Neither of the alternative labor supply assumptions substantially changes our results. To illustrate this we present estimates of the effect of H.R. 3734 on the number of persons in poverty under the three different labor supply assumptions (see Table 5). Alternative A represents our basic assumption, in which 56 percent of nonexempt AFDC mothers who reach the five-year time limit find part-time employment and 12 percent find full-time employment. Using this assumption, an estimated 2.6 million additional persons would fall into poverty, 1.1 million of whom are children. This represents a 9 percent increase in poverty for all persons and a 12 percent increase for children. Under the more pessimistic Alternative B, 30 percent would find part-time jobs and 10 percent full-time jobs. The number of persons in poverty would increase by 2.8 million (10 percent) and the number of children in poverty by 1.3 million (13 percent). Under the more optimistic Alternative C, 68 percent would find jobs (the same as Alternative A) but half would find full-time employment. The number of persons in poverty would increase by 2.5 million (9 percent) and the number of children in poverty by 1.1 million (11 percent). Some Possible Modifications to the Legislation The welfare reform legislation would make major changes to AFDC, SSI, and Food Stamps Programs at the same time. Some families depend on only one of these programs for support, but others receive benefits from several programs. The effects on family incomes and poverty obviously depend on the combined effects of all these provisions. Several proposals have been made to modify one or more provisions of the legislation in order to reduce the effects on family incomes but still retain the basic philosophy of the legislation. We show the effects of some of these options on poverty in Table 6.

1. Give states the option to continue some support for children whose parents have lost eligibility for welfare. Currently 31 states have approved or pending waivers to impose time limits. While half of these states would terminate all cash aid when the time limit is reached, others would either reduce aid or require participation in a work program when the time limit is reached. This option would allow states the flexibility to retain a minimum income floor for children and still receive federal matching funds to help cover the additional costs. The children's benefit (or voucher) would be equal to the child's portion of the AFDC grant for families who lose eligibility but do not find jobs. If all states implemented child vouchers, the number of children moved into poverty would be 140,000 fewer than under H.R.3734, according to our estimates. The increase in the poverty gap for families with children would be $700 million or 18 percent less than under H.R. 3734 ($5.2 billion compared to $5.9 billion).15 While this option would provide some financial support for families who reach the time limit and do not find jobs, it does not have a large effect on the number in poverty because most families on welfare are already in poverty even when their food stamp benefits are counted. 2. Increase the exemption from 20 to 25 percent of the caseload. This option would give states flexibility to exempt a larger portion of their caseloads from the time limits. It could be particularly useful in states where it is more difficult to move the welfare caseload into jobs because of relatively high unemployment rates or limited funds for job training and employment assistance. The effects of this option are similar to those shown for the child voucher. Both options would provide a way to protect more families and their children from the effects of a permanent time limit. 3. Restore noncitizen eligibility for SSI and Food Stamps. The proposed legislation would prohibit legal immigrants from receiving SSI or food stamp benefits. Some alternatives have been suggested that would allow current noncitizens to retain eligibility. Others would allow eligibility but increase the amount of sponsors' income counted in the income test for eligibility. To give some sense of the effects of these kinds of changes, we show the effects of simply restoring current rules applied to noncitizens. If eligibility for noncitizens were restored, 1.2 million fewer persons would be moved into poverty, including 450,000 children. The increase in the poverty gap would be narrowed to $3.4 billion, $2.5 billion less than under H.R. 3734. 4. Restore current law food stamp provisions for U.S. citizens. As described earlier, the welfare reform legislation would make many changes in the Food Stamp Program. Benefits for all families would be reduced, and nonworking, childless persons under age 51 would be limited to 3 months of benefits. These changes would reduce the size of the Food Stamp Program, with corresponding reductions in its capacity to offset the income losses from the other provisions of the legislation. Restoring current-law food stamp provisions for citizens would move 380,000 fewer children into poverty compared to H.R.3734. It would also dramatically reduce the effects of the legislation on the elderly and childless adult families. The increase in the poverty gap would be narrowed to $4.5 billion, $1.3 billion less than under H.R. 3734. Conclusions The welfare reform legislation recently passed by the House of Representatives, if it becomes law, is likely to reduce family incomes and increase poverty. One in ten families would have reductions in total income (including food stamps and other near cash benefits less taxes) averaging $1,300 per year. Based on this broad definition of income, the legislation would also move about 2.6 million more persons below the poverty line. These results hold even under optimistic assumptions about the ability of long-term welfare recipients to find jobs and under generous assumptions about states' choices under the new block grant legislation. The legislation attempts to decrease families' dependence on welfare by requiring work and strict time limits for welfare eligibility. At the same time it would eliminate SSI and food stamp benefits for legal immigrants, eliminate SSI eligibility for some children, and make substantial changes in the Food Stamp Program. When the legislation is fully phased in the combined provisions would reduce government assistance to low-income families by $16 billion per year compared to current law. There is broad consensus that the welfare system in this country needs to be reformed so that states have the flexibility to design more effective systems that move people from welfare to work. This legislation provides state flexibility and mandates strong work requirements for welfare recipients. At the same time it provides states fewer federal dollars and large reductions in other social welfare programs. The findings presented here suggest that the total package of changes would especially hurt low-income families with children. We show some possible modifications to the legislation that would retain more of the social safety net for low-income children and reduce the adverse effects on family incomes. The projected increase in poverty could be attenuated, for example, by increasing the hardship exemption, providing vouchers to support children whose parents do not find work at the end of the time limit, modifying the provisions that affect legal immigrants, and reducing the Food Stamp program cuts. Appendix A: Methods And Data16 Our analysis uses the Urban Institute's TRIM2 microsimulation model to calculate the effects of the proposed welfare legislation. The model contains a baseline description of each household, including detailed information about the amounts and sources of income. We estimate the effects of changes in social welfare spending programs by assuming the new program rules were in effect as alternatives to the current system. As described below we also make assumptions about how families will respond to new program rules to provide a range of possible results. The basic data for the microsimulation model are derived from the March 1994 Current Population Survey (CPS), with demographic information as of March 1994 and income information for calendar year 1993. The CPS is representative of the entire civilian, non-institutionalized population of the United States. We update these data to 1996 using historical changes in incomes and population size through 1994 and projections from the Census Bureau and the Congressional Budget Office for 1995 and 1996. To estimate the effects of changes in a tax or transfer program for each household, the model implements the same steps that a welfare caseworker would perform to determine benefits or that an H&R Block employee would follow to determine a client's tax liability. Our simulations rely primarily on 1993 program rules; however, we incorporate recently-enacted expansions of the EITC and Food Stamp Program that would be in effect as of 1996 in the absence of any legislative changes. The simulations are extremely detailed, capturing the actual rules and formulae of each tax and benefit program. In the benefit program simulations, eligibility and benefits vary by unit size and state of residence, among other variables. A unique strength of the detailed, microsimulation process is its ability to capture interactions among programs. For example, the TRIM2 model estimates the increase in food stamp benefits that would occur when other sources of income are reduced

Notes

1. The major provisions of H.R. 3734 are similar to those in the bill subsequently passed by the Senate, except for the food stamp provisions. The House bill differs from the Senate in that it includes an option for State Food Assistance block grants and more stringent restrictions on benefits for nonworking, childless adults. The two bills also differ in the changes to deductions and household definition provisions. CBO estimates that the House bill would reduce food stamp spending by $2.9 billion more than the Senate bill over the seven-year period. 2. The Urban Institute, December 11, 1995. Many of the changes proposed by H.R. 3734 were also included in the 1995 Budget Reconciliation Bill. 3. This includes state spending on AFDC and SSI, in addition to state general assistance programs. 4. These results are described more fully in "The Potential Effects of the Budget Reconciliation Act on Family Incomes," S. Zedlewski, S. Clark, E. Meier, and K. Watson, The Urban Institute, December 11, 1995. 5. The bill also includes a number of other provisions that would have a smaller effect on eligibility and benefit rules for the various programs. These changes are not modeled in this analysis. 6. The bill also provides an option for food assistance block grants in some states. This analysis makes the simplifying assumption that no states take this option. However, CBO estimates that this provision would reduce Food Stamp spending by $2.57 billion over the seven-year period. 7. The analysis assumes that child support collections for parents receiving IV-D services would be about $0.9 billion higher (in 1996 dollars) under these provisions. This estimate is based on an analysis of current child support awards and payment patterns in the Current Population Survey Child Support Supplement. 8. Alberto Martini, "How Important is it to Incorporate a Labor Supply Response when Simulating the Effects of Transfer Program Reforms?" The Urban Institute, April 1995. The analysis accounts for expected earnings less child care, other work expenses, and changes in benefits and taxes to estimate a response that would optimize single mothers' incomes. 9. The wage rate is based on an Urban Institute analysis of women represented in the 1990 Survey of Income and Program Participation. 10. For example, among women receiving AFDC for more than five years, 63 percent have less than a high school education and half have no prior work experience. See LaDonna Pavetti, "Who Is Affected by Time Limits?," in Welfare Reform: An Analysis of the Issues, Isabel V. Sawhill, editor, The Urban Institute, 1995. 11. Some argue that studies of the low-wage labor market indicate that many recipients reaching a time limit will seek but not find steady employment. See, for example, Sandra and Sheldon Danziger, "Will Welfare Recipients Find Work When Welfare Ends," in Welfare Reform: An Analysis of the Issues, Isabel V. Sawhill, editor, The Urban Institute, 1995. 12. See the 1995 OMB study entitled "Potential Poverty and Distributional Effects of Welfare Reform Bills and Balanced Budget Plans," November 9, 1995. 13. Some families would receive higher income under the welfare reform bill. For example, families entering the labor force may earn income greater than their previous AFDC benefit. Since Food Stamps fall by roughly $0.30 for every additional $1.00 in income, some additional savings may result from increased labor force participation among former welfare recipients. On the other hand, if a former AFDC recipient's income falls under the welfare reform bill, Food Stamp benefits may rise to offset a portion of the difference. The Food Stamp estimate captures these offsetting changes. 14. Using the official Census definition of poverty, which counts only cash income, the number of poor persons would increase by 0.8 million people under the proposed legislation — from 39.1 million to 39.9 million. 15. We show the effects of extending vouchers to children in families who have lost eligibility for welfare in all states because it is not possible to predict which states will take this option. This is consistent with the rest of our analysis that assumes that states implement the maximum allowable benefits. 16. A more detailed discussion of the model and data used for this analysis is presented in Zedlewski et al, (1995), op. cit.

Tables

Notes: H.R. 3734 includes a number of other provisions that would have a smaller effect on eligiblity and benefit rules for the various H.R. 3734 includes a number of other provisions that would have a smaller effect on eligiblity and benefit rules for the various programs. These changes are not modeled in this analysis.

1. This analysis assumes that all states exempt 20 percent of their caseloads.

2. This analysis assumes that states will not categorically deny benefits to noncitizens, although some noncitizens will lose eligibility due to deeming of sponsor income and the five-year prohibition.

3. This analysis assumes that states that do not currently have a "family cap" rule will opt out of this requirement, so that there is no net effect of this provision on family incomes.

4. States that qualify may elect to receive a State Food Assistance Block Grant in lieu of participation in the national Food Stamp Program, providing states with flexibility to define eligibility and benefits. This analysis assumes that all states decline the block grant and continue to participate in the national Food Stamp Program. Source: The Urban Institute's TRIM2 model, based on the March 1994 CPS aged to 1996.

Note: 1. Income is defined as after-tax income plus the value of food stamps, child nutrition programs, energy and housing subsidies, and the EITC.

2. Income losses of $100 or less per year are considered as no change in income.

Source: The Urban Institute's TRIM2 model, based on the March 1994 CPS aged to 1996.

Notes: 1. Income is defined as after-tax income plus the value of food stamps, child nutrition programs, energy and housing subsidies and the EITC.

2. Income losses of $100 or less per year are considered no change in income.

3. Family income and employment status are measured before the changes from H.R.3734.

4. Families are considered as working if annual earnings are greater than $1,000. Cash welfare includes AFDC and SSI.

Source: The Urban Institute's TRIM2 model, based on the March 1994 CPS aged to 1996.

Notes: 1. Income is defined as after-tax income plus the value of food stamps, child nutrition programs, energy and housing subsidies, and the EITC. 2. Labor supply alternatives apply to nonexempt, long-term recipients as follows: Alternative A: 56 percent find part-time jobs; 12 percent work full time Alternative B: 30 percent find part-time jobs; 10 percent work full time Alternative C: 34 percent find part-time jobs; 34 percent work full time Source: The Urban Institute's TRIM2 model, based on the March 1994 CPS aged to 1996. Notes: 1. Income is defined as after-tax income plus the value of food stamps, child nutritioin programs, energy and housing subsidies, and the EITC.

2. Alternatives to H.R. 3734 increase the maximum exemption from time limits; provide vouchers for children whose parents lose eligibility and do not find jobs; restore SSI and Food Stamp eligibility for legal immigrants; and restore current law food stamps provisions for U.S. citizens.

Other Publications by the Authors Sheila R. Zedlewski Sandra J. Clark Eric Meier Keith Watson

Usage and reprints: Most publications may be downloaded free of charge from the web site and may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact [email protected]. If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687. Disclaimer: The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute.